U.S. patent application number 11/244665 was filed with the patent office on 2006-04-06 for method for polarization-based intrusion monitoring in fiberoptic links.
Invention is credited to Gregory G. MacDonald, James J. JR. Sluss.
Application Number | 20060072922 11/244665 |
Document ID | / |
Family ID | 36125676 |
Filed Date | 2006-04-06 |
United States Patent
Application |
20060072922 |
Kind Code |
A1 |
MacDonald; Gregory G. ; et
al. |
April 6, 2006 |
Method for polarization-based intrusion monitoring in fiberoptic
links
Abstract
A fiber optic communication system including a fiber optic link,
a transmitter system and a receiver system. The transmitter system
includes a laser source producing a light beam, and a polarization
controller receiving the light beam and providing an expected
pattern of changing states of polarization to the light beam to
output light signals into the fiber optic link to cause the
expected pattern of changing states of polarization to be
transmitted along the fiber optic link. The receiver system is
provided with a polarization analyzer, and a light detector. The
light detector receives the light signals transmitted by the
transmitter, and forwards data indicative of the light signals to
the polarization analyzer. The polarization analyzer analyzing the
data with an inverse polarization reference frame and generates an
alert based on deviations of the data from the expected pattern of
changing states of polarization.
Inventors: |
MacDonald; Gregory G.;
(Owasso, OK) ; Sluss; James J. JR.; (Broken Arrow,
OK) |
Correspondence
Address: |
DUNLAP, CODDING & ROGERS P.C.
PO BOX 16370
OKLAHOMA CITY
OK
73113
US
|
Family ID: |
36125676 |
Appl. No.: |
11/244665 |
Filed: |
October 6, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60616328 |
Oct 6, 2004 |
|
|
|
Current U.S.
Class: |
398/152 |
Current CPC
Class: |
H04B 10/85 20130101 |
Class at
Publication: |
398/152 |
International
Class: |
H04B 10/00 20060101
H04B010/00 |
Claims
1. A fiber optic communication system, comprising: a fiber optic
link; a transmitter system comprising: a laser source producing a
light beam; and a polarization controller receiving the light beam
and providing an expected pattern of changing states of
polarization to the light beam to output light signals into the
fiber optic link to cause the expected pattern of changing states
of polarization to be transmitted along the fiber optic link; a
receiver system comprising: a polarization analyzer; and a light
detector receiving the light signals transmitted by the
transmitter, and forwarding data indicative of the light signals to
the polarization analyzer, the polarization analyzer analyzing the
data with an inverse polarization reference frame and generating an
alert based on deviations of the data from the expected pattern of
changing states of polarization.
2. The fiber optic communication system of claim 1, wherein the
inverse polarization reference frame is periodically recalibrated
to account for a slow time-varying birefringence component of the
fiber optic link.
3. The fiber optic communication system of claim 1, wherein the
polarization controller embeds calibration symbols in the expected
pattern of changing states of polarization.
4. The fiber optic communication system of claim 3, wherein the
polarization analyzer receives data indicative of the embedded
calibration symbols, and uses the embedded calibration symbols to
recalibrate the inverse polarization reference frame.
5. A method for polarization-based intrusion monitoring in a fiber
optic link, comprising the steps of: transmitting a light signal
having an expected pattern of changing states of polarization along
the fiber optic link; receiving the light signal and generating
data indicative of the light signal; and analyzing the data with an
inverse polarization reference frame; generating an alert based on
deviations of the data from the expected pattern of changing states
of polarization.
6. The method of claim 5, further comprising the step of
periodically recalibrating the inverse polarization reference frame
to account for a slow time-varying birefringence component of the
fiber optic link.
7. The method of claim 5, wherein the step of transmitting the
light signal further comprising the step of embedding multiple
calibration symbols in the expected pattern of changing states of
polarization to construct a calibration sequence.
8. The method of claim 7, wherein the polarization analyzer
receives data indicative of the embedded calibration symbols, and
uses the embedded calibration symbols to recalibrate the inverse
polarization reference frame.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The present patent application claims priority to the
provisional patent application identified by U.S. Serial No.
60/616,328 which was filed on Oct. 6, 2004. The entire content of
the provisional patent application identified by U.S. Serial No.
60/616,328 is hereby incorporated herein by reference.
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
[0002] Not Applicable.
BACKGROUND OF THE INVENTION
[0003] Optical communications is a fast advancing technology. As
the transmission of information using fiberoptic cables increases,
security risks become of growing concern, especially if military or
other highly sensitive information is being transmitted. One prior
art method of security for optical communications includes manual
surveillance. Another prior art method includes installing
fiberoptic cables in pressurized pipes, and then generating an
alert whenever pressure monitors detect a variation in the pressure
beyond some threshold (i.e., whenever a pipe is disturbed). These
methods can be inaccurate and generally require added human
resources and/or equipment in addition to that needed for
installation and operation of the optical network.
[0004] Information can be encoded for transmission using amplitude,
frequency, and polarization of a lightwave. While the first two
have been used for decades, the use of polarization to encode
information is relatively recent. Polarization in optical networks
has been used for multiplexing, modulation (polarization
shift-keying), cryptographic key generation, and as a data
encryption tool.
[0005] The present invention reduces such external needs and
provides a more efficient and effective method of providing
enhanced security for fiberoptic communication links and networks
by exploiting the inherent physical properties of fiberoptic cables
to detect physical "disturbances" in an optical network. More
particularly, the present invention relates to using state of
polarization information and the birefringence properties of fiber
optics to detect and monitor for physical intrusions of fiberoptic
based networks.
SUMMARY OF THE PRESENT INVENTION
[0006] The present invention relates to a method for
polarization-based intrusion monitoring in a fiber optic link. In
general, the method of the present invention utilizes polarization
information in a signal transmitted along a fiber optic link to
enhance the physical security of optical networks. In one
embodiment, an expected pattern of changing states of polarization
(SOPs) are generated and transmitted along the fiber optic link,
for example by using polarization-shift keying (POLSK). A deviation
from the expected pattern at the receiver system is interpreted as
an indication of a physical "disturbance" of the fiber.
[0007] In order to detect deviations from the expected pattern of
SOPs, the method of the present invention exploits the effects
caused by fiber birefringence. Birefringence is a term used to
describe the effects of refractive index asymmetry in optical
fibers. This asymmetry causes two orthogonal states of polarization
to propagate at different velocities through the fiber. The two
primary causes of fiber birefringence are fiber geometry and
stress. There is also a slow time-varying birefringence component
that is considered to be random and unpredictable. The end result
is that the original state of polarization at launch is transformed
as it propagates through the fiber.
[0008] A polarization reference frame can be established to model
the expected polarization transformation function of the fiber,
which is indicative of the effect on the state of polarization due
to birefringence caused by the geometry of the fiber, any current
stresses on the fiber, and the current effect of the time-varying
birefringence component. Once the polarization reference frame is
established, known SOPs can be launched and then measured at the
receiver system. The inverse of the polarization reference frame
can be applied to the measured SOPs received so as to derive
calculated-launched SOPs. If the calculated-launched SOPs are
different from the known-launched SOPs, i.e. if there are
deviations from the expected pattern of SOPs, then the deviations
can be assumed to be attributable to new stresses or "disturbances"
on the fiber. Thus, deviations from the expected pattern of SOPs
can be used to trigger the generation of an alert of a possible
"physical intrusion" or tampering of the optical network so as to
offer enhanced security for the optical network.
[0009] However, to more accurately identify deviations as being
caused by physical disturbances, the slow time-varying
birefringence component should also be compensated for. In the
present invention, the slow time-varying birefringence component is
compensated for by utilizing subsequently launched known SOPs to
determine a new polarization reference frame, thereby recalibrating
the polarization reference frame to account for changes in the
expected polarization transformation function of the fiber due to
the changed slow time-varying birefringence component. Also,
recalibrating the polarization reference frame allows the system to
adjust for new stresses on the fiber that will not be immediately
removed, such as for example when a new stress is caused by the
network owner or is not considered a security threat. Then the
system can continue transmitting data, and monitor for subsequent
new stresses on the fiber.
[0010] To detect possible physical disturbances and to compensate
for deviations caused by the relatively slow ("undisturbed")
time-varying nature of the birefringent properties of the fiber,
one preferred embodiment of the present invention uses a
"byte/symbol-stuffing" protocol to embed calibration symbols in a
stream of changing SOPs. An example of a method for
polarization-based intrusion monitoring in fiberoptic links which
uses the "byte/symbol-stuffing protocol" is as follows.
[0011] First, a polarization reference frame for the fiberoptic
link is established. As discussed above, the polarization reference
frame is the expected polarization transformation function
associated with the fiberoptic link, and generally depends on the
birefringence properties of the fiberoptic link caused by the
geometry of the fiberoptic link, any current stresses on the fiber
optic link, and the current effect of the time-varying
birefringence component. Determining the polarization reference
frame allows the effects of the fiber birefringence to be removed
at the receiver system.
[0012] The polarization reference frame can be established by
transmitting three known states of polarization (SOPs) along the
fiberoptic link, and then measuring each received SOP. The three
known launched SOPs and the three measured SOPs are then be used to
model the expected polarization transformation, such as for example
by determining a Jones matrix (for highly polarized light) or a
Mueller matrix (for partially polarized light). Although the
present invention has been described above as utilizing three known
launched SOPs to determine the polarization reference frame, it
should be understood that other numbers of known SOPs may be used
in accordance with the present invention, and will generally depend
on the method used to model the polarization transformation.
[0013] Once the polarization reference frame is established, known
SOPs are subsequently launched in an expanded data stream in the
form of three calibration symbols or "stuffed byte/symbols"
indicative of three known SOPs. In one embodiment, at least one of
the "stuffed bytes" is represented by a symbol that is different
than any symbols previously defined to represent general data in
the expanded data stream. Again, it should be understood that while
the present invention is described as using three calibration
symbols, a different number of calibration symbols can be used in
accordance with the present invention.
[0014] Polarization shift keying modulation is preferably used to
generate and transmit the expanded data stream. However, it should
be understood that other polarization data transfer techniques can
be used in accordance with the present invention.
[0015] When the data stream is received at the receiver system, the
SOPs corresponding to the transmitted calibration symbols are
measured. Calculated-launched SOPs are derived from the
measured-received SOPs by using the inverse of the polarization
reference frame to essentially remove the expected birefringence
effects of the fiberoptic link. The calculated-launched SOPs are
then compared to the known-launched SOPs to determine whether
deviations from the expected pattern of SOPs exist. If deviations
exist, an alert is generated indicating a possible physical
intrusion of the fiberoptic link.
[0016] Further, the polarization reference frame is recalibrated.
The calibration symbols of the transmitted expanded data stream
(i.e., the three known-launched SOPs and the three
measured-received SOPs) are utilized to determine a new or
"revised" polarization reference frame in the manner describe
above. The new polarization reference frame can then be used when
the above described steps are repeated for a subsequently
transmitted expanded data stream. As discussed above, by
continually or periodically revising or "updating" the polarization
transformation function of the fiberoptic link, the slow
time-varying birefringence component is tracked and better
compensated for. Also, the step of recalibrating the polarization
reference frame allows adaptation for desired or tolerated
environmental changes in the fiberoptic link so as to provide for
continuing transmission of data and monitoring.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] So that the above recited features and advantages of the
present invention can be understood in detail, a more particular
description of the invention, briefly summarized above, may be had
by reference to the embodiments thereof that are illustrated in the
appended drawings. It is to be noted, however, that the appended
drawings illustrate only typical embodiments of this invention and
are therefore not to be considered limiting of its scope, for the
invention may admit to other equally effective embodiments.
[0018] FIG. 1 illustrates an electric field vector E as the sum of
two orthogonal components E.sub.0x and E.sub.0y on a pair of
Cartesian coordinates X and Y.
[0019] FIG. 2 illustrates four stokes parameters mapped onto a
Poincare Sphere.
[0020] FIGS. 3(a-c) illustrate various states of linearly polarized
light and their corresponding Stokes parameters.
[0021] FIGS. 3(d-f) illustrates various states of right handed
elliptically polarized light and their corresponding Stokes
parameters.
[0022] FIG. 4 shows how a Jones matrix J is determined for a device
under test. More specifically, three knows polarization states
(Jones vectors) are produced at the transmitter and then measured
at the receiver system.
[0023] FIG. 5 illustrates a technique that involves dividing the
Poincare Sphere into a number of functional zones.
[0024] FIG. 6 is a screen shot of a computer illustrating a current
state of polarization along with the corresponding Stokes
parameters for point M.
[0025] FIG. 7A is a block diagram of a fiber optic communication
system constructed in accordance with the present invention.
[0026] FIG. 7B is another block diagram of the fiber optic
communication system.
[0027] FIG. 8 illustrates three matrices obtained during three
successive calibration sequences for a stationary fiber optic
link.
DETAILED DESCRIPTION OF THE INVENTION
[0028] Presently preferred embodiments of the invention are shown
in the above-identified figures and described in detail below. In
describing the preferred embodiments, like or identical reference
numerals are used to identify common or similar elements. The
figures are not necessarily to scale and certain features and
certain views of the figures may be shown exaggerated in scale or
in schematic in the interest of clarity and conciseness.
I. Introduction
[0029] Referring now to the drawings and in particular to FIG. 7A,
a fiber optic communication system 10 constructed in accordance
with the present invention is shown. The fiber optic communication
system 10 uses the polarization of the light signal to enhance the
physical security of one or more fiber optic link 12 in optical
network(s). In a preferred embodiment, a polarization-shift keying
(POLSK) scheme is used to generate an expected pattern of changing
states of polarization (SOPs). A deviation from the expected
pattern at the receiver system is due either to some physical
"disturbance" of the fiber, or the relatively slow ("undisturbed")
time-varying nature of the birefringent properties of the fiber
optic link. This last component is addressed using a
"byte/symbol-stuffing" protocol that embeds calibration symbols in
the stream of changing SOPs.
[0030] The technique described is applicable to any point-to-point
link where SOP fluctuations due to fiber birefringence can be
described by a rigid rotation of the launched SOP (i.e., rigid
rotation of the signal constellation points on the Poincare
sphere).
[0031] The fiber optic communication system 10 is also provided
with a transmitter system 14, and a receiver system 16. The
transmitter system 14 includes a laser source 20 producing a light
beam 21, and a polarization controller 22 receiving the light beam
21 and providing an expected pattern of changing states of
polarization to the light beam 21 to output light signals into the
fiber optic link 12 to cause the expected pattern of changing
states of polarization to be transmitted along the fiber optic link
12. The receiver system 16 is provided with a polarization analyzer
24, and a light detector 26. The light detector 26 receives the
light signals transmitted by the laser source 20, and forwards data
indicative of the light signals to the polarization analyzer 24.
The polarization analyzer 24 analyzes the data with an inverse
polarization reference frame and generates an alert based on
deviations of the data from the expected pattern of changing states
of polarization.
[0032] In a preferred embodiment, an initial calibration sequence
is required to establish the polarization reference frame. Then the
predetermined pattern begins and repeats. The predetermined pattern
includes "special" symbols that are "recognized" as re-calibration
symbols. In a sense, the special symbols are part of the
predetermined pattern in that if some disturbance occurs, the
expected special symbol that is to be used for recalibration will
not be received. In this case, what is received is still unexpected
and alert is raised.
[0033] In a commercial application, the polarization controller 22
would be incorporated into the transmitter system 14 and desirably
controlled by a local microprocessor 30, even though a remote
microprocessor could be used. The polarization analyzer 24 would be
incorporated into the receiver system 16 and would also desirably
operate under the control of a local microprocessor 32, which is
separate from the microprocessor controlling the polarization
controller 22.
[0034] An example of a suitable polarization controller is a model
HP 8160A obtainable from Hewlett Packard Co. In this type of
polarization controller, the light beam is `conditioned` to create
linearly polarized light, then it passes through two optical lens
to produce a particular state of polarization. The desired
polarized state is produced by rotating the two optical lens
relative to each other. The optical components are also known as
"retarders", e.g., quarter wavelength "retarders" and half
wavelength "restarders."
[0035] The patent application is organized as follows. Section II
is a brief review of the theoretical foundation for this work. This
section contains a brief review of polarization theory and a
description of polarization in terms of the Stokes parameters and
the Poincare sphere. Section III discusses two methods commonly
used to model the transformation of the SOP as the lightwave
propagates through single mode fiber (Jones Calculus and Mueller
Calculus). Section IV describes the 256-POLSK technique and the
design of the signal constellation pattern used for this work.
Section V describes the proposed "byte/symbol-stuffing" protocol
for polarization tracking and explains how it facilitates the
removal of the slow time-varying birefringent component. Section VI
discusses results for "undisturbed" fiber optic link 12 and
"disturbed" fiber optic link 12. Section VII concludes with a
discussion regarding the practical application of this
technique.
II. Theoretical Foundation
[0036] Electromagnetic radiation consists of oscillating electric
and magnetic fields. The electric field E and magnetic field H
vectors are mutually perpendicular and are both orthogonal to the
direction of propagation S (also called the Poynting vector). In
polarization, only the direction of the electric field vector E is
typically discussed.
[0037] Borrowing from Fowles [FOW75]: assuming cartesian
coordinates, and the direction of propagation S is along the
positive z-axis (towards the reader), the electric field E vector
(for a quasi-monochromatic source) can be described as the sum of
two orthogonal components (FIG. 1): E.sub.0= E.sub.0x+i
E.sub.0y
[0038] The corresponding wave function is: E=E.sub.0 exp i(kz-wt)
where, k is the propagation constant, z is the distance in the
direction of propagation, and w is the angular frequency.
[0039] If E.sub.0 is real, the electric field E vector represents
linearly polarized light. If E.sub.0 is complex the electric field
E vector represents elliptically polarized light. If the real and
imaginary parts of E.sub.0 are equal the electric field E vector
represents circularly polarized light. The relative amplitudes and
phases of the two orthogonal components determine the direction and
path traced out by the tip of the electric field E over time (i.e.,
the resulting state of polarization). If the direction of rotation
is clockwise, the polarization state is said to be right-handed. If
the direction of rotation is counterclockwise, the polarization
state is said to be left-handed.
[0040] Although the above description of polarization is in terms
of the electric field of the lightwave, it is generally not
convenient to measure polarization in such a manner. Fortunately,
it is possible to measure the optical power of the lightwave and
derive its state of polarization. One common method of expressing
the state of polarization is in terms of the Stokes parameters.
Stokes parameters are determined by a set of intensity measurements
taken when light is passed through various types of polarizers.
Stokes parameters are often mapped onto the Poincare sphere (FIG.
2). Combined, the Stokes parameters offer a way to conveniently
represent any state of polarization.
[0041] The value of each parameter is based on measured intensity
levels. The four Stokes parameters are defined as [GRE93]:
S.sub.0=total power(polarized+unpolarized)
S.sub.1=S.sub.0*cos(2.gamma.)*cos(2.beta.) (eq.1)
S.sub.2=S.sub.0*cos(2.gamma.)*sin(2.beta.) (eq.2)
S3=S.sub.0*sin(2.gamma.) (eq.3) The physical interpretation of the
above is as follows: [0042] S.sub.0=Total power of the received
signal (polarized+unpolarized) [0043] S.sub.1=Power received
through a horizontal linear polarizer-Power received through a
vertical linear polarizer [0044] S.sub.2=Power received through a
45 degree linear polarizer-Power receive through a-45 degree linear
polarizer [0045] S.sub.3=Power received through right hand circular
polarizer-Power received through a left hand circular polarizer The
"normalized" Stokes parameters are given by: s1=S.sub.1/S.sub.0
s2=S.sub.2/S.sub.0 s3=S.sub.3/S.sub.0 where for fully polarized
light,
S.sub.0={(S.sub.1).sup.2+(S.sub.2).sup.2+(S.sub.3)}.sup.1/2
[0046] Each point on the sphere represents a particular SOP. The
region of the sphere where S.sub.3=0 ("equator") describes various
orientations of linearly polarized light (see FIGS. 3a, 3b, and
3c). Areas where S.sub.3>0 ("northern hemisphere") represent
right handed elliptically polarized light (see FIGS. 3d, 3e, and
3f), while areas where S.sub.3<0 ("southern hemisphere")
represent left handed elliptically polarized light.
[0047] The eccentricity of the ellipse depends on its "latitude"
while the orientation of the ellipse depends on the "longitude". As
we move from the "equator" to the "poles" the polarization
ellipticity decreases from 1 to 0. The upper most point on the
sphere ("north pole") represents right handed circularly polarized
light, and the lower most point on the sphere ("south pole")
represents left handed circularly polarized light. It should be
noted that antipodal points on the Poincare sphere represent
mutually orthogonal states of polarization.
III. Fiber Birefringence and the Polarization Transfer Function
[0048] Birefringence is a term used to describe the effects of
refractive index asymmetry in optical fibers. This asymmetry causes
two orthogonal states of polarization to propagate at different
velocities through the fiber optic link 12. The two primary causes
of fiber birefringence are the geometry and stress of the fiber
optic link 12. There is also a slow time-varying birefringence
component that is considered to be random and unpredictable. The
end result is that the original state of polarization at launch is
transformed as it propagates through the fiber optic link 12.
[0049] From previous work the following is known:
[0050] If two mutually orthogonal light signals are launched into a
fiber, they emerge at the receiver with their orthogonality
preserved [CIM87].
[0051] The time-varying nature of the state of polarization at the
receiver ranges from seconds (during installation) to hours (for
already installed fiber) [HAR82].
[0052] Evolution of the SOP along single mode fiber can be
described as a rigid rotation of points (SOPs) on the Poincare
sphere (i.e., described by the Jones Calculus for fully polarized
light, or the Mueller Calculus for partially polarized light)
[SIM77].
[0053] The Jones Calculus yields a 2.times.2 matrix that describes
the polarization transformation function in terms of amplitudes and
phases. In general, the elements of the Jones matrix are complex.
The Mueller Calculus yields a 4.times.4 matrix that describes the
polarization transfer function in terms of intensity measurements.
The elements of the Mueller matrix are real. It should be noted
that the Mueller matrix can be derived from the Jones matrix. The
relationship is a function of the elements of the Jones matrix and
their complex conjugates [GER94]. Since the data in the appendices
(B and C) show we are working with "fully" polarized light (DOP
column), we will discuss only the Jones Calculus in detail.
[0054] The Jones method depends on optical field measurements as
opposed to signal intensity measurements. Fully polarized light can
be described by a two-element complex vector (sometimes called the
Jones Vector) of the form: E = ( E X E Y ) ##EQU1##
[0055] For example, the Jones vector for linearly polarized light
is: E = ( 1 0 ) ##EQU2## while the Jones vector for right-handed
circularly polarized light is: E = 1 SQRT .function. ( 2 ) .times.
( 1 + i ) ##EQU3##
[0056] The Jones matrix for any device can be determined by
measuring the resulting output Jones vectors for three known input
Jones vectors [DER98]. While any three distinct input Jones vectors
can be used, in practice it is common to use three linear SOPs as
shown in FIG. 4.
[0057] The relationship between the launch SOP(l) and the received
SOP(r) then is: E(r)=J*E(l) where J is the 2.times.2 Jones matrix.
The launched SOP(l) can be recovered from the received SOP(r) by:
E(l)=J.sup.-1*E(r) (eq.4)
IV. 256-Polarization Shift-Keying
[0058] A number of authors have studied polarization shift keying
modulation techniques [BET90][BEN95][BEN97]. The general theory of
polarization shift-keying (2-POLSK, 4-POLSK, and 8-POLSK systems)
is given by Benedetto [BEN92]. This section describes a simple
256-POLSK technique that involves dividing the Poincare sphere into
a number of functional zones.
[0059] The Poincare sphere is divided into a number of functional
zones, e.g., 256 areas consisting of eight operational zones and
one calibration zone (see table 1, and FIGS. 5 and 6). The purpose
is to produce 256 different states of polarization, each SOP
representing 1 of the 256 different bit combinations (hexadecimal
00 through FF). Each operational zone is divided into a particular
number of segments based on the circumference of the zone center.
For example, zone 1 is "centered" at S.sub.3=0.8 and zone 2 is
centered at S.sub.3=0.6. Since the radius of zone 2 is greater than
the radius of zone 1, zone 2 will be divided into a greater number
of areas representing a larger number of polarized states (see
appendix A for the division for zone 4). Therefore, zones having
greater radii have greater resolvable symbol capacity.
TABLE-US-00001 TABLE 1 Hex Hex Zone Zone Polarized Byte Zone Zone
Polarized Byte Number Function S.sub.3 Range States Value Number
Function S.sub.3 Range States Value 1 operational 0.7 < S.sub.3
< 0.9 23 69-7F 1 operational 0.7 < S.sub.3 < 0.9 23 69-7F
2 operational 0.5 < S.sub.3 < 0.7 31 4A-68 2 operational 0.5
< S.sub.3 < 0.7 31 4A-68 3 operational 0.3 < S.sub.3 <
0.5 35 27-49 3 operational 0.3 < S.sub.3 < 0.5 35 27-49 4
operational 0.1 < S.sub.3 < 0.3 38 01-26 4 operational 0.1
< S.sub.3 < 0.3 38 01-26 5 calibration -0.1 < S.sub.3 <
+0.1 2 00 and 80 5 calibration -0.1 < S.sub.3 < +0.1 2 00 and
80 6 operational -0.3 < S.sub.3 < -0.1 38 81-A6 6 operational
-0.3 < S.sub.3 < -0.1 38 81-A6 7 operational -0.5 <
S.sub.3 < -0.3 35 A7-C9 7 operational -0.5 < S.sub.3 <
-0.3 35 A7-C9 8 operational -0.7 < S.sub.3 < -0.5 31 CA-EA 8
operational -0.7 < S.sub.3 < -0.5 31 CA-EA 9 operational -0.9
< S.sub.3 < -0.7 23 E8-FF 9 operational -0.9 < S.sub.3
< -0.7 23 E8-FF
Operational Zones and Their Corresponding Symbol Range
[0060] In one example of the present invention depicted in FIG. 7B,
the polarization controller 22 and the polarization analyzer 24 are
controlled by a personal computer 40. The desired states of
polarization and resulting measurements were achieved using the
setup shown in FIG. 7B. The personal computer 40 contains a GPIB
card 42 and connects to the polarization analyzer 24 and the
polarization controller 22 by way of a GPIB bus 44. The fiber optic
link 12 was formed by three meters of single mode fiber. In this
exemplary embodiment, the controlling entity is a C program
(Illustrated in Appendix D of Exhibit A of the provisional patent
application identified by U.S. Serial No. 60/616,328) running on
the personal computer 40. The program issues commands to the
devices across the GPIB bus 44 and uses the same bus 44 to retrieve
data measured by the polarization analyzer 24. Devices are
controlled using vendor supplied device drivers and a library of C
subroutines.
[0061] The C program begins by establishing a polarization
reference frame (i.e., measuring the Jones matrix), so the effects
of the fiber birefringence (polarization transfer function) can be
determined and later removed during the measurement process. This
procedure sends instructions to the polarization controller 22 to
produce three known states of polarization. The program then
instructs the polarization analyzer 24 to measure each received
SOP. These three measurements are sufficient to acquire the Jones
matrix for the device under test (i.e., the single mode fiber).
Once these measurements are complete, subsequently launched SOPs
can be derived from the received SOPs using the relationship
defined in eq.4. (Note: the inverse transformation is performed by
the analyzer once the polarization reference frame is
established).
[0062] The program maps 256 positions (signal constellation
points), for example, on the Poincare sphere as described above.
Each position is described in terms of angles .beta. and .gamma.
(FIG. 2). The appropriate Stokes parameters for each .beta. and
.gamma. are calculated using the equations given in section II
(eqs. 1-3). The program instructs the polarization controller 22 to
produce a unique SOP for each symbol. It then instructs the
polarization analyzer 24 to take 100 measurements, returning only
the average values for S.sub.1, S.sub.2, and S.sub.3. A "reverse"
lookup is performed by the program in order to recover the received
symbol. Mid-points of each interval are used as symbol decision
criteria. A subset of results for "undisturbed" fiber optic link 12
for zone 4 is given in Appendix B of the provisional patent
application identified by U.S. Serial No. 60/616,328.
V. A Protocol for Polarization Tracking
[0063] Algorithms for tracking and recovering the random drift of
signal constellation points on the Poincare sphere due to fiber
birefringence have been proposed and studied [BEN97]. The following
description provides a protocol designed to track and compensate
for the slow time-varying nature of the changing SOP. We take
advantage of the fact that this slow rate of SOP fluctuation is
much slower than the bit-rate [NIC89].
[0064] We define hexadecimal symbol "00" to be represented by
Stokes parameters S.sub.1=1, S.sub.2=0, S.sub.3=0; and hexadecimal
"80" to be represented by Stokes parameters S.sub.1=-1, S.sub.2=0,
S.sub.3=0. These two symbols reside in the calibration zone (FIG.
5). The appearance of these symbols in the SOP stream will be
followed by three calibration symbols ("stuffed byte/symbols")
having the following Stokes parameters (points A, B, and C in FIG.
6): his figure shows zone 4 and part of zone 3. The red lines are
on the front side of the sphere while the blue lines are on the
back side of the sphere. The display on the left side of this
figure shows the current state of polarization along with the
corresponding Stokes parameters for point M. Points A, B, and C are
the calibration points (next section). S1=1.0 S2=0.0, S3=0 (first
"stuffed byte") S1=-0.500, S2=0.866, S3=0 (second "stuffed symbol")
S1=-0.500, S2=-0.866, S3=0 (third "stuffed symbol")
[0065] It should be noted the second and third "stuffed bytes" are
not any of the 256 previously defined symbols. Instead, they are
simply two additional SOPs. All "stuffed" entities represent linear
SOPs.
Consider the following raw SOP stream:
[0066] 1F 20 4D 55 00 3F EF 5F 75 B3 80 EE This stream would be
expanded to include the calibration symbols: ##STR1##
[0067] The "stuffed" byte/symbols represent three known states of
polarization, making the determination of a new Jones matrix
(J.sub.i) possible. As explained in Section III, this "revised"
Jones matrix describes the evolving polarization transfer function.
This essentially removes the slow time-varying birefringence
component. Any remaining deviations from the expected SOP are
likely to be due to some physical disturbance of the fiber optic
link 12. In the C program, if the expected SOP is not recovered, a
"physical intrusion" alert is generated and then the system is
recalibrated.
VI. Discussion of Experimental Results
[0068] Appendices A, B, C and D in Exhibit A of the provisional
patent application identified by U.S. Serial No. 60/616,328 are
hereby expressly incorporated herein by reference in their
entirety. Appendices A, B, and C contain data relevant to the
256-POLSK technique described above and more specifically describe
examples of the invention, the production thereof and uses
thereof.
[0069] Appendix A shows how zone 4 can be partitioned into 38
different segments. Each segment is defined by a set of Stokes
parameters (S.sub.1, S.sub.2, S.sub.3). Each set of Stokes
parameters describes a unique state of polarization, and hence a
different symbol. The values shown for S.sub.1, S.sub.2, and
S.sub.3 represent the "center" values. The angle column represents
a point on the Poincare sphere rotated counterclockwise about the
S.sub.3 axis (as viewed from above the sphere). Appendix A contains
the theoretical Stokes parameters required to produce the desired
SOPs for operational zone 4. The graphs for other zones will be
similar with the exception of a vertical shift in the S.sub.3 line
and a different number of points for the S.sub.1 and S.sub.2
parameters (reflecting a different radius for each zone).
[0070] Appendix B shows the Stokes parameter values specified to
produce different SOPs and the received Stokes parameter
measurements. Appendix B represents the fiber optic link 12 that is
"undisturbed". The specified and measured values track very
closely, reflecting little variation in the evolving Jones matrix.
Appendix B shows a table illustrating a subset (zone 4) of the
experimentally obtained results. The table in Appendix B shows the
symbol to be sent, the required Stokes parameters (state of
polarization) to produce the symbol [S(s)], the measured Stokes
parameters at the receiver system [S(r)], and the recovered symbol.
The data presented in this section is typical for the fiber optic
link 12 that is "undisturbed." The "DOP" column is the degree of
polarization.
[0071] Although not related to the data presented in Appendix B,
Jones matrix measurements for the "undisturbed" fiber optic link 12
typically evolve slowly as shown in FIG. 8.
[0072] Appendix C shows the received Stokes parameters for the
"disturbed" fiber optic link 12 case. In this case, the fiber
disturbance at symbol 28 (hex 1C) is clearly evident and results in
the recovery of symbol 3D. Since this is not the expected symbol,
recalibration is performed resulting in a "revised" Jones matrix.
Subsequent data appear to track nicely with the expected SOP
sequence only because the revised Jones matrix has been applied.
Although not shown, any attempt to place the fiber optic link 12
back into its undisturbed position would result in another
unexpected symbol, and a recalibration. Since not all regions of
the Poincare sphere contained defined symbols, disturbed fiber
optic link 12 may produce Stokes parameters that do not represent a
defined SOP. All that is required to generate an alert is the
recovery of an unexpected SOP or a set of Stokes parameters that do
not represent a defined SOP. The table in Appendix C shows the
symbol to be sent, the required Stokes parameters (state of
polarization) to produce the symbol [S(s)], the measured Stokes
parameters at the receiver system [S(r)], and the recovered symbol.
The data presented in this section includes randomly "disturbed"
fiber optic link 12 at symbol "1C". Again, "DOP" column is the
degree of polarization.
VII. Final Remarks
[0073] An existing technique used by the U.S. government to monitor
network cables for tampering activities involve installing fiber in
pressurized pipes. Alerts are generated whenever pressure monitors
detect a variation in the pressure that deviates from some
threshold. The assumption here is pressure readings that exceed
threshold values are caused by some physical disturbance of the
pressurized pipes.
[0074] The present invention is directed to an alternate technique
that uses polarization and the birefringent properties of a fiber
optic link 12, e.g., single mode fiber, to provide enhanced
protection for optical networks. A received SOP that differs from
the expected SOP is due either to a physical disturbance of the
fiber optic link 12, or the slow time-varying nature of the
properties of the fiber optic link 12. The latter is tracked using
a "byte/symbol stuffing" protocol. A new Jones matrix is determined
from these calibration symbols. Recalibration occurs whenever the
data stream contains these "special" SOPs and whenever the received
SOP differs from the expected SOP. By taking advantage of the fact
that the slow time-varying component occurs much slower than the
bit-rate, we can compensate for this component. Received SOP's that
differ from the expected SOP's are presumed to be caused by a
physical disturbance somewhere along the fiber optic link 12.
Alerts are issued indicating a potential tampering situation.
[0075] It will be understood from the foregoing description that
various modifications and changes may be made in the preferred and
alternative embodiments of the present invention without departing
from its true spirit. For example, if the tracking is so far off
that we can no longer recognize the calibration sequence, this
becomes an unexpected deviation and the polarization reference
frame is recalibrated. Further, although a system using 256 symbols
has been described, it is reasonable to use a constellation pattern
that having far fewer symbols. In fact, it is possible to conceive
of an adaptive technique that varies the number of symbols in use
over time. Intuitively, a system using fewer constellation signal
points would probably be more tolerant of a wider variation of
environmental conditions, while a system using an adaptive approach
would be more responsive to changing conditions.
[0076] How frequently should the SOP be changed? Although we varied
the SOP each bit period, our bit periods were relatively long in
time. The speed of the GPIB bus 44 and a firmware problem with the
polarization analyzer 24 required us to deliberately introduce
elongated bit periods. Another approach might involve maintaining
the same SOP over many bit periods. In this instance, the SOP
period is independent of the bit period. Rapid SOP measurements may
also be possible. A recent study [VAN99] demonstrated the ability
to perform SOP measurements at the nanosecond level.
[0077] How does an out of order SOP effect the bit error rate? In
this technique polarization is essentially used as a separate
channel. As such, we do not use the SOP to bear actual data bit
information. As long as the receiving equipment is insensitive to
fluctuations in polarization, out of order SOPs should not impair
data bit flow.
[0078] Appendix D contains an exemplary C program used to map the
constellation points on the Poincare sphere, control the
polarization controller 22, and gather the data from the
polarization analyzer 24.
[0079] However, although the foregoing invention has been described
in some detail by way of illustration and example for purposes of
clarity of understanding, it will be apparent to those skilled in
the art that certain changes and modifications may be practiced
without departing from the spirit and scope thereof, as described
herein and in the above-referenced attachments.
REFERENCES
[0080] 1. The contents of each of the below identified references
are hereby incorporated herein by reference in their entirety.
[0081] 2. Benedetto, S., Poggiolini, P., 1992. Theory of
Polarization Shift Keying Modulation. IEEE Transactions on
Communications. Vol. 40, No. 4. Apr. 4, 1992. [BEN92] [0082] 3.
Benedetto S., Gaudino, R., Poggiolini, P., 1997. Polarization
Recovery in Optical Polarization Shift-Keying Systems. IEEE
Transactions on Communications. Vol. 45, No. 10, October, 1997.
[BEN97] [0083] 4. Benedetto S., Olmo R., Poggiolini P., 1995.
Trellis Coded Polarization Shift Keying Modulation for Digital
Optical Communications. IEEE Transactions on Communications. Vol.
43, No. 2, pp. 1591-1602, February 1995. [BEN95] [0084] 5. Betti
S., Curti, F., De Marchis, G., Iannone E., 1990. Multilevel
Coherent Optical System Based on Stokes Parameters Modulation.
Journal of Lightwave Technology, Vol. 8, No. 7, July 1990. [BET90]
[0085] 6. Cimini, L. J., Habbab, I. M., John, R. K., Saleh, A. A,
1987. On the Preservation of Polarization Orthogonality a Through
Linear System. Electronic Letters. Vol. 23, No. 23, pp. 1365-1366,
Dec. 3, 1987. [CIM87] [0086] 7. Derickson, D., 1998. Fiber Optic
Test and Measurement. Princeton-Hall, Inc. 1998. [DER98] [0087] 8.
Fowles, Grant, R., 1975. Introduction to Modern Optics. Dover
Publications, Inc., 2nd Ed., [FOW75] [0088] 9. Gerrard, A., Burch,
J. M., 1994. Introduction to Matrix Methods in Optics. Dover
Publications, Inc. (Originally published in 1975). [GER94] [0089]
10. Green, Paul, E., 1993. Fiber Optic Networks. Princeton Hall
PTR. [GRE93] [0090] 11. Harmon, R. A., 1982. Polarization Stability
in Long Lengths of Monomode Fiber. Electronic Letters. Vol. 18, No.
24, pp. 1058-1060, November 1982. [HAR82] [0091] 12. Nicholson, G.,
Temple, D. J., 1989. Polarization Fluctuation Measurement on
Installed Single-Mode Optical Fiber Cables. Journal of Lightwave
Technology. Vol. 7, August, 1989. [NIC89] [0092] 13. Simon, A.,
Ulrich, R., 1977. Evolution of Polarization along a Single Mode
Fibre. Applied Physics Letters. Vol. 31, No. 8, pp. 517-520, 1977.
[SIM77] [0093] 14. Van Wiggeren, G., Roy, R., 1999. High-Speed
Fiber-Optic Polarization Analyzer: Measurements of the Polarization
Dynamics of an Erbium-doped Fiber Ring Laser. Optical
Communications. Issue 164, pp. 107-120, 1999. [VAN99] [0094] 15.
Van Wiggeren, G., Roy, R., 2002. Communication with Dynamically
Fluctuating States of Light Polarization. Physical Review Letters.
Vol 88., No 9. Mar. 4, 2002. [VAN02]
* * * * *